evorAI
The Behavioural Science and AI tool
evorAI - AI-powered behavioural insight for better decisions, services and products
Organisations increasingly collect valuable free-text data, from open-text survey responses and interview transcripts to consultation feedback, public online discussions and other free-text sources.
But analysing this data consistently can be time-consuming and resource-intensive. Teams may have rich insights at their fingertips, but limited time, capacity or specialist resources to turn them into clear, practical recommendations or actionable insights.
evorAI is designed to help bridge this gap.
What evorAI helps with
evorAI supports teams to move from “what people said” to clearer insights about patterns, needs, experiences, opportunities and possible next steps.
It analyses free-text data through a behavioural science lens, helping teams identify meaningful patterns and turn insight into useful outputs for communication, engagement, product development, service improvement, programmes and digital tools.
The aim is not to replace expert judgement, but to make behavioural insight work faster, more structured and more transparent.
How evorAI works
Many AI tools can summarise text. evorAI goes further by applying a behavioural science lens. This means the analysis is not limited to identifying themes. It also considers what those themes may tell us about practical constraints and other factors that shape behaviour.
Add your text data
Prepare text-based data for analysis, including survey responses, interview extracts, social media comments, consultation feedback, service-user feedback or other free-text data.
Step 1Let evorAI map the data
evorAI maps the data against behavioural science frameworks and related behaviour change concepts.
Step 2Explore the patterns
The dashboard presents mapped data in a clear visual format, helping users explore patterns, compare categories and identify areas for further analysis.
Step 3Turn insight into action
evorAI helps users explore mapped data, identify patterns and create draft outputs, while human oversight ensures outputs are reviewed and used appropriately.
Step 4Example case study
Client: Holland & Barrett
Project focus: Mapping barriers and facilitators of behaviour change
Context: A 12-week evaluation of a Holland & Barrett health and wellbeing app
Data analysed: Free-text reflections from staff using the app.
Responses analysed: 609
Behavioural focus areas: Physical activity, nutrition, sleep and emotional wellbeing.
Purpose of analysis: To explore how evorAI can support faster, structured analysis of text-based insight while retaining expert interpretation and quality assurance.
Analytical approach: Comparison of manual behavioural analysis with evorAI-supported analysis involving human validation.
evorAI delivered major efficiency gains.
With human validation only, estimated analysis time fell to 2 hours, compared with 41 hours for a junior behavioural scientist approach and 28 hours for an experienced behavioural scientist approach — making evorAI 20.5× faster and 14× faster, respectively.
Why this mattered
The case study illustrates how evorAI can support organisations to analyse text-based data more efficiently, while retaining behavioural science expertise, human interpretation and quality assurance.
Quotes from the usability testing
-
“At the moment, we’restabbing in the dark.
This tool shows us exactlywhat the barriers are, so wecan focus our messageswhere they’ll make adifference.”
-
“This is incredible. In anemergency situation, youhaven’t got time to do interviewsor surveys, this tool helps youquickly understand what peopleare saying and tailor messagesaccordingly.”
-
“It’s like doing weeks ofmanual thematic analysisin a click. The timesavings alone make itworth it.”
-
“It’s fast, evidence-based, and built on realbehavioural science,that’s what makes it feelcredible and useful.”
Why us?
evorAI is grounded in Applied Behaviour Change’s expertise in behavioural science, public health, digital health, evaluation and intervention development.
Our work combines academic rigour with practical experience of designing, evaluating and improving services in real-world settings. We understand that behavioural insight needs to be useful, proportionate and actionable for the teams who will use it.
evorAI builds on Innovate UK-supported development work and ongoing research collaboration, including current work with UCL through Behavioural Research UK. The tool is also informed by a multidisciplinary consortium spanning public health, artificial intelligence, behavioural science, policy and applied research.
This helps ensure evorAI is scientifically grounded, responsibly developed and useful for real-world decision-making.
Consortium Members
-
Prof Katherine Brown
Public Health Expert
The University of Hertfordshire
-
Prof Andy Pardoe
AI expert, Chair of the Deep Tech Innovation Centre at the University of Warwick. Founder & Chief AI Officer of Informed AI.
-
Katie Wilson
Health Improvement Manager, Warwickshire County Council
-
Dr Chryssa Stefanidou
Lead Behavioural/Social Scientist
Office for Health Improvement and Disparities - DHSC
-
Dr Lucy Porter
Behaviour Science Expert
Centre for Behaviour Change, University College London
-
Dr Paulina Bondaronek
HUMBLE project lead (Human-centred Method for Bias-reducing Algorithms with NLP and Qualitative analysis for Improved Health Outcomes of Underserved and Marginalised Populations)
University College London
Interested in using evorAI?
We are currently interested in speaking with public health teams, local authorities, charities, researchers, digital health teams, health and wellbeing organisations, consumer health brands, retailers and service providers who want to explore how AI-assisted behavioural insight could support their work.
Contact us to discuss:
Pilot projects
Research collaborations
Consultation or engagement analysis
Behavioural insight reports
AI and behavioural science partnerships.